EXCEED THE SPACE PROVIDED. WARCEF-STAT and WARCEF-CLIN are two separate but highly coordinated units which together constitute WARCEF, a two-arm (1:1) double-blind randomized multicenter clinical trial (target enrollment 2860 patients at 70 clinical sites) designed to test the primary null hypothesis of no difference between warfarin (INR 2.5-3, target 2.75) and aspirin (325 mg/day) in 3-5 year survival for the composite endpoint death or stroke or intracerebral hemorrhage among patients with low (<=30%) cardiac Left Ventricular Ejection Fraction (LVEF). Prespecified secondary hypotheses are tested for subgroups of women and African Americans, and for the endpoint of stroke alone. Prespecified tertiary hypotheses are tested by amalgamating data from WARCEF and another trial (WATCH). The goals of WARCEF-STAT are: 1) to conduct, efficiently integrate, and be responsible for all WARCEF data management, double-blinding, medication management, quality control, and reporting operations. 2) to specify, conduct and be responsible for all interim, safety, and final statistical analyses for WARCEF and for the reporting of these results. 3) to transmit to the Clinical Coordinating Center (CCC) in a timely fashion all of the data which the principal investigator of WARCEF-CLIN needs to meet his responsibilities in terms of ensuring patient safety and directing all clinical operations. Goal 1 is accomplished by integrating a Data Management Center with substantial experience in cardiac trials and an anticoagulation unit which is successfully conducting an ongoing double blind warfarin-aspirin study (WARSS). The double-blind algorithm used in WARSS to fabricate INRs for patients on active aspirin will be recalibrated and revalidated for the WARCEF target range before WARCEF begins. Goal 2 is achieved by a statistical analysis unit that is experienced in stroke, cardiac, and other clinical trials. Goal 3 is ensured by an Operations Committee of WARCEF-STAT and WARCEF-CLIN investigators and senior coordinators which meets frequently under tightly specified and controlled communication protocols.